Using client-side event logging and path tracing to assess and improve the quality of web-based surveys.
- 1 January 2002
- journal article
- p. 894-8
Abstract
Web-based data collection has considerable appeal. However, the quality of data collected using such instruments is often questionable. There can be systematic problems with the wording of the surveys, and/or the means with which they are deployed. In unsupervised data collection, there are also concerns about whether subjects understand the questions, and wehther they are answering honestly. This paper presents a schema for using client-side timestamps and traces of subjects' paths through instruments to detect problems with the definition of instruments and their deployment. We discuss two large, anonymous, web-based, medical surveys as examples of the utility of this approach.This publication has 9 references indexed in Scilit:
- A comparison of World-Wide Web and paper-and-pencil personality questionnairesBehavior Research Methods, Instruments & Computers, 2002
- Randomized Testing of Alternative Survey Formats Using Anonymous Volunteers on the World Wide WebJournal of the American Medical Informatics Association, 2001
- Web Survey Design and AdministrationPublic Opinion Quarterly, 2001
- The capture and use of detailed process information in the Dialogix system for structured web-based interactions.2001
- Web surveys - A review of issues and approaches2000
- Methods for the Design and Administration of Web-based SurveysJournal of the American Medical Informatics Association, 2000
- Longitudinal Analysis of the Effects of the Aging Process on Neuropsychological Test Performance in the Healthy Young-Old and Oldest-OldDevelopmental Neuropsychology, 2000
- Same information, different decisions: format countsBMJ, 1999
- A self-assessment questionnaire to determine morningness-eveningness in human circadian rhythms.1976